February 1, 2020

3151 words 15 mins read

Paper Group AWR 334

GraphTSNE: A Visualization Technique for Graph-Structured Data. Neural Embedding Propagation on Heterogeneous Networks. Attention Based Pruning for Shift Networks. DANE: Domain Adaptive Network Embedding. Fast Visual Object Tracking with Rotated Bounding Boxes. SNU_IDS at SemEval-2019 Task 3: Addressing Training-Test Class Distribution Mismatch in …

GraphTSNE: A Visualization Technique for Graph-Structured Data


Title	GraphTSNE: A Visualization Technique for Graph-Structured Data
Authors	Yao Yang Leow, Thomas Laurent, Xavier Bresson
Abstract	We present GraphTSNE, a novel visualization technique for graph-structured data based on t-SNE. The growing interest in graph-structured data increases the importance of gaining human insight into such datasets by means of visualization. Among the most popular visualization techniques, classical t-SNE is not suitable on such datasets because it has no mechanism to make use of information from the graph structure. On the other hand, visualization techniques which operate on graphs, such as Laplacian Eigenmaps and tsNET, have no mechanism to make use of information from node features. Our proposed method GraphTSNE produces visualizations which account for both graph structure and node features. It is based on scalable and unsupervised training of a graph convolutional network on a modified t-SNE loss. By assembling a suite of evaluation metrics, we demonstrate that our method produces desirable visualizations on three benchmark datasets.
Tasks	Dimensionality Reduction
Published	2019-04-15
URL	http://arxiv.org/abs/1904.06915v3
PDF	http://arxiv.org/pdf/1904.06915v3.pdf
PWC	https://paperswithcode.com/paper/graphtsne-a-visualization-technique-for-graph
Repo	https://github.com/leowyy/GraphTSNE
Framework	pytorch

Neural Embedding Propagation on Heterogeneous Networks


Title	Neural Embedding Propagation on Heterogeneous Networks
Authors	Carl Yang, Jieyu Zhang, Jiawei Han
Abstract	Classification is one of the most important problems in machine learning. To address label scarcity, semi-supervised learning (SSL) has been intensively studied over the past two decades, which mainly leverages data affinity modeled by networks. Label propagation (LP), however, as the most popular SSL technique, mostly only works on homogeneous networks with single-typed simple interactions. In this work, we focus on the more general and powerful heterogeneous networks, which accommodate multi-typed objects and links, and thus endure multi-typed complex interactions. Specifically, we propose \textit{neural embedding propagation} (NEP), which leverages distributed embeddings to represent objects and dynamically composed modular networks to model their complex interactions. While generalizing LP as a simple instance, NEP is far more powerful in its natural awareness of different types of objects and links, and the ability to automatically capture their important interaction patterns. Further, we develop a series of efficient training strategies for NEP, leading to its easy deployment on real-world heterogeneous networks with millions of objects. With extensive experiments on three datasets, we comprehensively demonstrate the effectiveness, efficiency, and robustness of NEP compared with state-of-the-art network embedding and SSL algorithms.
Tasks	Network Embedding
Published	2019-09-29
URL	https://arxiv.org/abs/1910.00005v1
PDF	https://arxiv.org/pdf/1910.00005v1.pdf
PWC	https://paperswithcode.com/paper/neural-embedding-propagation-on-heterogeneous
Repo	https://github.com/JieyuZ2/NEP
Framework	pytorch

Attention Based Pruning for Shift Networks


Title	Attention Based Pruning for Shift Networks
Authors	Ghouthi Boukli Hacene, Carlos Lassance, Vincent Gripon, Matthieu Courbariaux, Yoshua Bengio
Abstract	In many application domains such as computer vision, Convolutional Layers (CLs) are key to the accuracy of deep learning methods. However, it is often required to assemble a large number of CLs, each containing thousands of parameters, in order to reach state-of-the-art accuracy, thus resulting in complex and demanding systems that are poorly fitted to resource-limited devices. Recently, methods have been proposed to replace the generic convolution operator by the combination of a shift operation and a simpler 1x1 convolution. The resulting block, called Shift Layer (SL), is an efficient alternative to CLs in the sense it allows to reach similar accuracies on various tasks with faster computations and fewer parameters. In this contribution, we introduce Shift Attention Layers (SALs), which extend SLs by using an attention mechanism that learns which shifts are the best at the same time the network function is trained. We demonstrate SALs are able to outperform vanilla SLs (and CLs) on various object recognition benchmarks while significantly reducing the number of float operations and parameters for the inference.
Tasks	Object Recognition
Published	2019-05-29
URL	https://arxiv.org/abs/1905.12300v1
PDF	https://arxiv.org/pdf/1905.12300v1.pdf
PWC	https://paperswithcode.com/paper/attention-based-pruning-for-shift-networks
Repo	https://github.com/eghouti/SAL
Framework	pytorch

DANE: Domain Adaptive Network Embedding


Title	DANE: Domain Adaptive Network Embedding
Authors	Yizhou Zhang, Guojie Song, Lun Du, Shuwen Yang, Yilun Jin
Abstract	Recent works reveal that network embedding techniques enable many machine learning models to handle diverse downstream tasks on graph structured data. However, as previous methods usually focus on learning embeddings for a single network, they can not learn representations transferable on multiple networks. Hence, it is important to design a network embedding algorithm that supports downstream model transferring on different networks, known as domain adaptation. In this paper, we propose a novel Domain Adaptive Network Embedding framework, which applies graph convolutional network to learn transferable embeddings. In DANE, nodes from multiple networks are encoded to vectors via a shared set of learnable parameters so that the vectors share an aligned embedding space. The distribution of embeddings on different networks are further aligned by adversarial learning regularization. In addition, DANE’s advantage in learning transferable network embedding can be guaranteed theoretically. Extensive experiments reflect that the proposed framework outperforms other state-of-the-art network embedding baselines in cross-network domain adaptation tasks.
Tasks	Domain Adaptation, Network Embedding
Published	2019-06-03
URL	https://arxiv.org/abs/1906.00684v2
PDF	https://arxiv.org/pdf/1906.00684v2.pdf
PWC	https://paperswithcode.com/paper/190600684
Repo	https://github.com/lukashedegaard/dage
Framework	tf

Fast Visual Object Tracking with Rotated Bounding Boxes


Title	Fast Visual Object Tracking with Rotated Bounding Boxes
Authors	Bao Xin Chen, John K. Tsotsos
Abstract	In this paper, we demonstrate a novel algorithm that uses ellipse fitting to estimate the bounding box rotation angle and size with the segmentation(mask) on the target for online and real-time visual object tracking. Our method, SiamMask_E, improves the bounding box fitting procedure of the state-of-the-art object tracking algorithm SiamMask and still retains a fast-tracking frame rate (80 fps) on a system equipped with GPU (GeForce GTX 1080 Ti or higher). We tested our approach on the visual object tracking datasets (VOT2016, VOT2018, and VOT2019) that were labeled with rotated bounding boxes. By comparing with the original SiamMask, we achieved an improved Accuracy of 0.652 and 0.309 EAO on VOT2019, which is 0.056 and 0.026 higher than the original SiamMask. The implementation is available on GitHub: https://github.com/baoxinchen/siammask_e.
Tasks	Object Tracking, Visual Object Tracking
Published	2019-07-08
URL	https://arxiv.org/abs/1907.03892v5
PDF	https://arxiv.org/pdf/1907.03892v5.pdf
PWC	https://paperswithcode.com/paper/fast-visual-object-tracking-with-rotated
Repo	https://github.com/baoxinchen/siammask_e
Framework	pytorch

SNU_IDS at SemEval-2019 Task 3: Addressing Training-Test Class Distribution Mismatch in Conversational Classification


Title	SNU_IDS at SemEval-2019 Task 3: Addressing Training-Test Class Distribution Mismatch in Conversational Classification
Authors	Sanghwan Bae, Jihun Choi, Sang-goo Lee
Abstract	We present several techniques to tackle the mismatch in class distributions between training and test data in the Contextual Emotion Detection task of SemEval 2019, by extending the existing methods for class imbalance problem. Reducing the distance between the distribution of prediction and ground truth, they consistently show positive effects on the performance. Also we propose a novel neural architecture which utilizes representation of overall context as well as of each utterance. The combination of the methods and the models achieved micro F1 score of about 0.766 on the final evaluation.
Tasks
Published	2019-03-06
URL	http://arxiv.org/abs/1903.02163v2
PDF	http://arxiv.org/pdf/1903.02163v2.pdf
PWC	https://paperswithcode.com/paper/snu_ids-at-semeval-2019-task-3-addressing
Repo	https://github.com/baaesh/semeval19_task3
Framework	pytorch

On the Smoothness of Nonlinear System Identification


Title	On the Smoothness of Nonlinear System Identification
Authors	Antônio H. Ribeiro, Koen Tiels, Jack Umenberger, Thomas B. Schön, Luis A. Aguirre
Abstract	New light is shed onto optimization problems resulting from prediction error parameter estimation of linear and nonlinear systems. It is shown that the smoothness” of the objective function depends both on the simulation length and on the decay rate of the prediction model. More precisely, for regions of the parameter space where the model is not contractive, the Lipschitz constant and $\beta$-smoothness of the objective function might blow up exponentially with the simulation length, making it hard to numerically find minima within those regions or, even, to escape from them. In addition to providing theoretical understanding of this problem, this paper also proposes the use of multiple shooting as a viable solution. The proposed method minimizes the error between a prediction model and observed values. Rather than running the prediction model over the entire dataset, as in the original prediction error formulation, multiple shooting splits the data into smaller subsets and runs the prediction model over each subdivision, making the simulation length a design parameter and making it possible to solve problems that would be infeasible using a standard approach. The equivalence with the original problem is obtained by including constraints in the optimization. The method is illustrated for the parameter estimation of nonlinear systems with chaotic or unstable behavior, as well as on neural network parameter estimation.
Tasks
Published	2019-05-02
URL	https://arxiv.org/abs/1905.00820v1
PDF	https://arxiv.org/pdf/1905.00820v1.pdf
PWC	https://paperswithcode.com/paper/on-the-smoothness-of-nonlinear-system
Repo	https://github.com/antonior92/MultipleShootingPEM.jl
Framework	none

Reusing Convolutional Activations from Frame to Frame to Speed up Training and Inference


Title	Reusing Convolutional Activations from Frame to Frame to Speed up Training and Inference
Authors	Arno Khachatourian
Abstract	When processing similar frames in succession, we can take advantage of the locality of the convolution operation to reevaluate only portions of the image that changed from the previous frame. By saving the output of a layer of convolutions and calculating the change from frame to frame, we can reuse previous activations and save computational resources that would otherwise be wasted recalculating convolutions whose outputs we have already observed. This technique can be applied to many domains, such as processing videos from stationary video cameras, studying the effects of occluding or distorting sections of images, applying convolution to multiple frames of audio or time series data, or playing Atari games. Furthermore, this technique can be applied to speed up both training and inference.
Tasks	Atari Games, Time Series
Published	2019-09-02
URL	https://arxiv.org/abs/1909.05632v2
PDF	https://arxiv.org/pdf/1909.05632v2.pdf
PWC	https://paperswithcode.com/paper/reusing-convolutional-activations-from-frame
Repo	https://github.com/arnokha/reusing_convolutions
Framework	pytorch

Potential of deep features for opinion-unaware, distortion-unaware, no-reference image quality assessment


Title	Potential of deep features for opinion-unaware, distortion-unaware, no-reference image quality assessment
Authors	Subhayan Mukherjee, Giuseppe Valenzise, Irene Cheng
Abstract	Image Quality Assessment algorithms predict a quality score for a pristine or distorted input image, such that it correlates with human opinion. Traditional methods required a non-distorted “reference” version of the input image to compare with, in order to predict this score. However, recent “No-reference” methods circumvent this requirement by modelling the distribution of clean image features, thereby making them more suitable for practical use. However, majority of such methods either use hand-crafted features or require training on human opinion scores (supervised learning), which are difficult to obtain and standardise. We explore the possibility of using deep features instead, particularly, the encoded (bottleneck) feature maps of a Convolutional Autoencoder neural network architecture. Also, we do not train the network on subjective scores (unsupervised learning). The primary requirements for an IQA method are monotonic increase in predicted scores with increasing degree of input image distortion, and consistent ranking of images with the same distortion type and content, but different distortion levels. Quantitative experiments using the Pearson, Kendall and Spearman correlation scores on a diverse set of images show that our proposed method meets the above requirements better than the state-of-art method (which uses hand-crafted features) for three types of distortions: blurring, noise and compression artefacts. This demonstrates the potential for future research in this relatively unexplored sub-area within IQA.
Tasks	Image Quality Assessment, No-Reference Image Quality Assessment
Published	2019-11-27
URL	https://arxiv.org/abs/1911.11903v1
PDF	https://arxiv.org/pdf/1911.11903v1.pdf
PWC	https://paperswithcode.com/paper/potential-of-deep-features-for-opinion
Repo	https://github.com/subhayanmukherjee/deepiqa
Framework	tf

Towards a complete 3D morphable model of the human head


Title	Towards a complete 3D morphable model of the human head
Authors	Stylianos Ploumpis, Evangelos Ververas, Eimear O’ Sullivan, Stylianos Moschoglou, Haoyang Wang, Nick Pears, William A. P. Smith, Baris Gecer, Stefanos Zafeiriou
Abstract	Three-dimensional Morphable Models (3DMMs) are powerful statistical tools for representing the 3D shapes and textures of an object class. Here we present the most complete 3DMM of the human head to date that includes face, cranium, ears, eyes, teeth and tongue. To achieve this, we propose two methods for combining existing 3DMMs of different overlapping head parts: i. use a regressor to complete missing parts of one model using the other, ii. use the Gaussian Process framework to blend covariance matrices from multiple models. Thus we build a new combined face-and-head shape model that blends the variability and facial detail of an existing face model (the LSFM) with the full head modelling capability of an existing head model (the LYHM). Then we construct and fuse a highly-detailed ear model to extend the variation of the ear shape. Eye and eye region models are incorporated into the head model, along with basic models of the teeth, tongue and inner mouth cavity. The new model achieves state-of-the-art performance. We use our model to reconstruct full head representations from single, unconstrained images allowing us to parameterize craniofacial shape and texture, along with the ear shape, eye gaze and eye color.
Tasks
Published	2019-11-18
URL	https://arxiv.org/abs/1911.08008v2
PDF	https://arxiv.org/pdf/1911.08008v2.pdf
PWC	https://paperswithcode.com/paper/towards-a-complete-3d-morphable-model-of-the
Repo	https://github.com/steliosploumpis/Universal_Head_3DMM
Framework	none

Camera Exposure Control for Robust Robot Vision with Noise-Aware Image Quality Assessment


Title	Camera Exposure Control for Robust Robot Vision with Noise-Aware Image Quality Assessment
Authors	Ukcheol Shin, Jinsun Park, Gyumin Shim, Francois Rameau, In So Kweon
Abstract	In this paper, we propose a noise-aware exposure control algorithm for robust robot vision. Our method aims to capture the best-exposed image which can boost the performance of various computer vision and robotics tasks. For this purpose, we carefully design an image quality metric which captures complementary quality attributes and ensures light-weight computation. Specifically, our metric consists of a combination of image gradient, entropy, and noise metrics. The synergy of these measures allows preserving sharp edge and rich texture in the image while maintaining a low noise level. Using this novel metric, we propose a real-time and fully automatic exposure and gain control technique based on the Nelder-Mead method. To illustrate the effectiveness of our technique, a large set of experimental results demonstrates higher qualitative and quantitative performances when compared with conventional approaches.
Tasks	Image Quality Assessment
Published	2019-07-11
URL	https://arxiv.org/abs/1907.12646v1
PDF	https://arxiv.org/pdf/1907.12646v1.pdf
PWC	https://paperswithcode.com/paper/camera-exposure-control-for-robust-robot
Repo	https://github.com/WookCheolShin/Noise-AwareCameraExposureControl
Framework	none

Transfer Learning for Performance Modeling of Configurable Systems: A Causal Analysis


Title	Transfer Learning for Performance Modeling of Configurable Systems: A Causal Analysis
Authors	Mohammad Ali Javidian, Pooyan Jamshidi, Marco Valtorta
Abstract	Modern systems (e.g., deep neural networks, big data analytics, and compilers) are highly configurable, which means they expose different performance behavior under different configurations. The fundamental challenge is that one cannot simply measure all configurations due to the sheer size of the configuration space. Transfer learning has been used to reduce the measurement efforts by transferring knowledge about performance behavior of systems across environments. Previously, research has shown that statistical models are indeed transferable across environments. In this work, we investigate identifiability and transportability of causal effects and statistical relations in highly-configurable systems. Our causal analysis agrees with previous exploratory analysis \cite{Jamshidi17} and confirms that the causal effects of configuration options can be carried over across environments with high confidence. We expect that the ability to carry over causal relations will enable effective performance analysis of highly-configurable systems.
Tasks	Transfer Learning
Published	2019-02-26
URL	http://arxiv.org/abs/1902.10119v1
PDF	http://arxiv.org/pdf/1902.10119v1.pdf
PWC	https://paperswithcode.com/paper/transfer-learning-for-performance-modeling-of
Repo	https://github.com/majavid/AAAI-WHY-2019
Framework	none

Interpret Federated Learning with Shapley Values


Title	Interpret Federated Learning with Shapley Values
Authors	Guan Wang
Abstract	Federated Learning is introduced to protect privacy by distributing training data into multiple parties. Each party trains its own model and a meta-model is constructed from the sub models. In this way the details of the data are not disclosed in between each party. In this paper we investigate the model interpretation methods for Federated Learning, specifically on the measurement of feature importance of vertical Federated Learning where feature space of the data is divided into two parties, namely host and guest. For host party to interpret a single prediction of vertical Federated Learning model, the interpretation results, namely the feature importance, are very likely to reveal the protected data from guest party. We propose a method to balance the model interpretability and data privacy in vertical Federated Learning by using Shapley values to reveal detailed feature importance for host features and a unified importance value for federated guest features. Our experiments indicate robust and informative results for interpreting Federated Learning models.
Tasks	Feature Importance
Published	2019-05-11
URL	https://arxiv.org/abs/1905.04519v1
PDF	https://arxiv.org/pdf/1905.04519v1.pdf
PWC	https://paperswithcode.com/paper/interpret-federated-learning-with-shapley
Repo	https://github.com/crownpku/federated_shap
Framework	none

Evaluation of Retinal Image Quality Assessment Networks in Different Color-spaces


Title	Evaluation of Retinal Image Quality Assessment Networks in Different Color-spaces
Authors	Huazhu Fu, Boyang Wang, Jianbing Shen, Shanshan Cui, Yanwu Xu, Jiang Liu, Ling Shao
Abstract	Retinal image quality assessment (RIQA) is essential for controlling the quality of retinal imaging and guaranteeing the reliability of diagnoses by ophthalmologists or automated analysis systems. Existing RIQA methods focus on the RGB color-space and are developed based on small datasets with binary quality labels (i.e., `Accept' and` Reject’). In this paper, we first re-annotate an Eye-Quality (EyeQ) dataset with 28,792 retinal images from the EyePACS dataset, based on a three-level quality grading system (i.e., `Good',` Usable’ and `Reject’) for evaluating RIQA methods. Our RIQA dataset is characterized by its large-scale size, multi-level grading, and multi-modality. Then, we analyze the influences on RIQA of different color-spaces, and propose a simple yet efficient deep network, named Multiple Color-space Fusion Network (MCF-Net), which integrates the different color-space representations at both a feature-level and prediction-level to predict image quality grades. Experiments on our EyeQ dataset show that our MCF-Net obtains a state-of-the-art performance, outperforming the other deep learning methods. Furthermore, we also evaluate diabetic retinopathy (DR) detection methods on images of different quality, and demonstrate that the performances of automated diagnostic systems are highly dependent on image quality. \|
Tasks	Image Quality Assessment
Published	2019-07-10
URL	https://arxiv.org/abs/1907.05345v4
PDF	https://arxiv.org/pdf/1907.05345v4.pdf
PWC	https://paperswithcode.com/paper/evaluation-of-retinal-image-quality
Repo	https://github.com/hzfu/EyeQ
Framework	none

Disentangling Interpretable Generative Parameters of Random and Real-World Graphs


Title	Disentangling Interpretable Generative Parameters of Random and Real-World Graphs
Authors	Niklas Stoehr, Emine Yilmaz, Marc Brockschmidt, Jan Stuehmer
Abstract	While a wide range of interpretable generative procedures for graphs exist, matching observed graph topologies with such procedures and choices for its parameters remains an open problem. Devising generative models that closely reproduce real-world graphs requires domain knowledge and time-consuming simulation. While existing deep learning approaches rely on less manual modelling, they offer little interpretability. This work approaches graph generation (decoding) as the inverse of graph compression (encoding). We show that in a disentanglement-focused deep autoencoding framework, specifically Beta-Variational Autoencoders (Beta-VAE), choices of generative procedures and their parameters arise naturally in the latent space. Our model is capable of learning disentangled, interpretable latent variables that represent the generative parameters of procedurally generated random graphs and real-world graphs. The degree of disentanglement is quantitatively measured using the Mutual Information Gap (MIG). When training our Beta-VAE model on ER random graphs, its latent variables have a near one-to-one mapping to the ER random graph parameters n and p. We deploy the model to analyse the correlation between graph topology and node attributes measuring their mutual dependence without handpicking topological properties.
Tasks	Graph Embedding, Graph Generation, Graph Representation Learning
Published	2019-10-12
URL	https://arxiv.org/abs/1910.05639v2
PDF	https://arxiv.org/pdf/1910.05639v2.pdf
PWC	https://paperswithcode.com/paper/disentangling-interpretable-generative
Repo	https://github.com/niklasstoehr/thesis
Framework	none